Tumour Antigens, Status:
T cell defined antigens have now been characterised in a broad spectrum of cancer types. These antigens can be divided into several main groups, depending on their expression. The two main groups are constituted by developmental differentiation related antigens (tumour-testis antigens, oncofoetal antigens etc., such as MAGE antigens and CEA) and tissue specific differentiation antigens (Tyrosinase, gp100 etc.). The group containing the truly tumour specific antigens contains proteins that are altered due to mutations in the genes encoding them. In the majority of these, the mutations are unique and have been detected in a single or in a small number of tumours. Several of these antigens seem to play a role in oncogenesis.
Cancer Vaccines, Status:
The focus in cancer vaccine development has been on antigens expressed in a high degree within one form of cancer (such as melanoma) or in many kinds of cancers. One reason for this is the increased recruitment of patients into clinical protocols. The field is in rapid growth, illustrated by the accompanying table listing the cancer vaccine protocols currently registered in the PDQ database of NCI.
Inheritable Cancer/Cancer Gene Testing:
Inherited forms of cancer occur at a certain frequency in the population. For several of these cancer forms, the underlying genetic defects have been mapped. This is also the case in Lynch syndrome cancers which constitute an important group of inheritable cancer. In families inflicted with this syndrome, family members inherit defect genes encoding DNA Mismatch Repair (MMR) Enzymes. Carriers of such MMR defects frequently develop colorectal cancer (HNPCC) and other forms of cancer (list?). Mutations in MMR enzymes can be detected using gene testing in the same way as other cancer related genes can be detected.
Gene testing of risk groups in this case represents an ethical dilemma, since no acceptable forms for prophylactic treatment exist. At present surgery to remove the organ in danger to develop cancer has been the only treatment option. In these patients, cancer vaccines will be a very (interesting) form of prophylaxis, provided efficient vaccines can be developed.
The lack of efficient repair of mismatched DNA results in deletions and insertions in one strand of DNA, and this happens preferentially in stretches of DNA containing repeated units (repeat sequences). Until now, focus has been on repeat sequences in the form of non-coding microsattelite loci. Indeed microsattelite instability is the hallmark of cancers resulting from MMR defects. We have taken another approach, and have concentrated on frameshift mutations occurring in DNA sequences coding for proteins related to the oncogenic process. Such frameshift mutations result in completely new amino acid sequences in the C-terminal part of the proteins, prematurely terminating where a novel stop codon appears. This results in two important consequences:
1) The truncated protein resulting from the frameshift is generally nonfunctional, in most cases resulting in “knocking out” of an important cellular function. Aberrant proteins may also gain new functions such as the capacity to aggragate and form plaques. In both cases the frameshift results in disease.
2) The short new C-terminal amino acid sequence resulting from the shift in the reading frame (the “frameshift sequence”) is foreign to the body. It does not exist prior to the mutation, and it only exists in cells having the mutation, i.e. in tumour cells and their pre malignant progenitors. Since they are completely novel and therefore foreign to the immune system of the carrier, they may be recognised by T-cells in the repertoire of the carrier. So far, nobody has focused on this aspect of frameshift mutations, and no reports exist on the characterisation of frameshift peptides from coding regions of proteins as tumour antigens. This concept is therefore novel and forms the basis for developing vaccines based on these sequences. It follows that such vaccines may also be used prophyllactively in persons who inherit defective enzymes belonging to the MMR machinery. Such vaccines will therefore fill an empty space in the therapeutic armament against inherited forms of cancer.
It has been shown that single amino acid substitutions in intracellular “self”-proteins may give rise to tumour rejection antigens, consisting of peptides differing in their amino acid sequence from the normal peptide. The T cells which recognise these peptides in the context of the major histocompatibility (MHC) molecules on the surface of the tumour cells, are capable of killing the tumour cells and thus rejecting the tumour from the host.
In contrast to antibodies produced by the B cells, which typically recognise a free antigen in its native conformation and further potentially recognise almost any site exposed on the antigen surface, T cells recognise an antigen only if the antigen is bound and presented by a MHC molecule. Usually this binding will take place only after appropriate antigen processing, which comprises a proteolytic fragmentation of the protein, so that the resulting peptide fragment fits into the groove of the MHC molecule. Thereby T cells are enabled to also recognise peptides derived from intracellular proteins. T cells can thus recognise aberrant peptides derived from anywhere in the tumour cell, in the context of MHC molecules on the surface of the tumour cell, and can subsequently be activated to eliminate the tumour cell harbouring the aberrant peptide.
M. Barinaga, Science, 257, 880–881, 1992 offers a short review of how MHC binds peptides. A more comprehensive explanation of the Technical Background for this Invention may be found in D. Male et al, Advanced Immunology, 1987, J.B.lippincott Company, Philadelphia. Both references are hereby included in their entirety.
The MHC molecules in humans are normally referred to as HLA (human leukocyte antigen) molecules. They are encoded by the HLA region on the human chromosome No 6.
The HLA molecules appear as two distinct classes depending on which region of the chromosome they are encoded by and which T cell subpopulations they interact with and thereby activate primarily. The class I molecules are encoded by the HLA A, B and C subloci and they primarily activate CD8+ cytotoxic T cells. The HLA class II molecules are encoded by the DR, DP and DQ subloci and primarily activate CD4+ T cells, both helper cells and cytotoxic cells.
Normally every individual has six HLA Class I molecules, usually two from each of the three groups A, B and C. Correspondingly, all individuals have their own selection of HLA Class II molecules, again two from each of the three groups DP, DQ and DR. Each of the groups A, B, C and DP, DQ and DR are again divided into several subgroups. In some cases the number of different HLA Class I or II molecules is reduced due to the overlap of two HLA subgroups.
All the gene products are highly polymorphic. Different individuals thus express distinct HLA molecules that differ from those of other individuals. This is the basis for the difficulties in finding HLA matched organ donors in transplantations. The significance of the genetic variation of the HLA molecules in immunobiology is reflected by their role as immune-response genes. Through their peptide binding capacity, the presence or absence of certain HLA molecules governs the capacity of an individual to respond to peptide epitopes. As a consequence, HLA molecules determine resistance or susceptibility to disease.
T cells may control the development and growth of cancer by a variety of mechanisms. Cytotoxic T cells, both HLA class I restricted CD8+ and HLA Class II restricted CD4+, may directly kill tumour cells carrying the appropriate tumour antigens. CD4+ helper T cells are needed for cytotoxic CD8+ T cell responses as well as for antibody responses, and for inducing macrophage and LAK cell killing.
A requirement for both HLA class I and II binding is that the peptides must contain a binding motif, which usually is different for different HLA groups and subgroups. A binding motif is characterised by the requirement for amino acids of a certain type, for instance the ones carrying large and hydrophobic or positively charged side groups, in definite positions of the peptide so that a narrow fit with the pockets of the HLA binding groove is achieved. The result of this, taken together with the peptide length restriction of 8–10 amino acids within the binding groove, is that it is quite unlikely that a peptide binding to one type of HLA class I molecules will also bind to another type. Thus, for example, it may very well be that the peptide binding motif for the HLA-A1 and HLA-A2 subgroups, which both belong to the class I gender, are as different as the motifs for the HLA-A1 and HLA-B1 molecules.
For the same reasons it is not likely that exactly the same sequence of amino acids will be located in the binding groove of the different class II molecules. In the case of HLA class II molecules the binding sequences of peptides may be longer, and it has been found that they usually contain from 10 to 16 amino acids, some of which, at one or both terminals, are not a part of the binding motif for the HLA groove.
However, an overlap of the different peptide binding motifs of several HLA class I and class II molecules may occur. Peptides that have an overlap in the binding sequences for at least two different HLA molecules are said to contain “nested T cell epitopes”. The various epitopes contained in a “nested epitope peptide” may be formed by processing of the peptide by antigen presenting cells and thereafter be presented to T cells bound to different HLA molecules. The individual variety of HLA molecules in humans makes peptides containing nested epitopes more useful as general vaccines than peptides that are only capable of binding to one type of HLA molecule.
Effective vaccination of an individual can only be achieved if at least one type of HLA class I and/or II molecule in the patient can bind a vaccine peptide either in it's full length or as processed and trimmed by the patient's own antigen presenting cells.
The usefulness of a peptide as a general vaccine for the majority of the population increases with the number of different HLA molecules it can bind to, either in its full length or after processing by antigen presenting cells.
In order to use peptides derived from a protein encoded by a mutated gene as vaccines or anticancer agents to generate anti tumour CD4+ and/or CD8+ T cells, it is necessary to investigate the mutant protein in question and identify peptides that are capable, eventually after processing to shorter peptides by the antigene presenting cells, to stimulate T cells.